A few Ideas I'd like to lay down about making the improve system point based rather than chance (if it is chance based, which i'm certain it is from experience and in-game chats)

Reasons - 1. Fits well with the RP aspect: Instead of casting or using a skill and hoping that the percentage will tick, the more one uses the skill the more one gets better albeit things like intelligence, level of player and opponent, and how high is the skill/spell percentage affects the rate of progression.

2. Less of a pain for all classes that require a certain percentage in order to attain better stuff not just for invokers: Read once on qhcf.net that getting lore to 80% allows more properties of an item to be revealed like damroll and hitroll (had an elf warrior with 24 int and was still pain in the ass to raise lore skill up by one point, never got past 78% before deletion). Giant warriors (or any warrior for that matter) could benefit greatly since most lack Int. Assassins could raise their skills in kicks and technques quicker. Bards could reach that golden 100% sing skill to get all their songs (another trouble I had for a lvl 30ish bard that always stayed at 99%)any other skills ,spells ,etc. i did not include is because I haven't played that much with said skills, spells, etc.

Outline - All skill usages will gain a certain amount of points before increasing in which the amount is determined by the usual level of character and opponent(s), skill percentage, affinity for invokers, etc.

There are some unique features I'd like to elaborate on but I'd like to see how much support this idea first

I know there are plenty of characters out there who have been doing fine with the current Skill improvement but to me it seems more like a hassle to spam and hope for better skills.

I don't think I would prefer that system, the only real advantage I see to it is now botters would know exactly how many times to spam a spell in their scripts. I don't think that is what the imms are shooting for.

>A few Ideas I'd like to lay down about making the improve
>system point based rather than chance (if it is chance based,
>which i'm certain it is from experience and in-game chats)

I don't understand what you mean by this.

>>A few Ideas I'd like to lay down about making the improve
>>system point based rather than chance (if it is chance
>based,
>>which i'm certain it is from experience and in-game chats)
>
>I don't understand what you mean by this.

Lemme re-word that:

Based on in-game experience and question and answer moments, I found the skill improvement mechanic in CF was chance-based (the rate at which a skill improved was influenced by factors but ultimately wound up to how lucky one was).

I was proposing a leveling system for skills much like how a player levels in rank, except instead of levels, it's for a percentage point. The more one uses a skill, the more "xp" a skill gets and after a certain amount of "skill xp" the percentage increases.

>>>A few Ideas I'd like to lay down about making the improve
>>>system point based rather than chance (if it is chance
>>based,
>>>which i'm certain it is from experience and in-game chats)
>>
>>I don't understand what you mean by this.
>
>Lemme re-word that:
>
>Based on in-game experience and question and answer moments, I
>found the skill improvement mechanic in CF was chance-based
>(the rate at which a skill improved was influenced by factors
>but ultimately wound up to how lucky one was).
>
>I was proposing a leveling system for skills much like how a
>player levels in rank, except instead of levels, it's for a
>percentage point. The more one uses a skill, the more "xp" a
>skill gets and after a certain amount of "skill xp" the
>percentage increases.


This system wont change anything compared to how it is today. By the law of averages it will even out with the same amount of casts to master a spell. Under your system it would take a character with 25 int 200 casts to master a spell, and 1000 casts to master 5 spells. Under the current system mastering those 5 spells would also take 1000 casts, only difference is that spell 1 may take 250 casts and spell 2 may take 150 casts and so on.

Lets use an anology. If you flip a coin 10.000 times, the law of averages say that you would end up with approximatly 5000 tails and 5000 heads. It may lean slighly towards one end, but in the big picture its pretty close to 50-50. Under your system the thing that change is that you introduce a machine that can flip the coin exactly enaugh in a controlled enviroment, so that it lands on tails half the time and heads the other half. Which ends up with exactly 5000 tails and exactly 5000 heads. Essentially you just take the chance out of it, without actually changing anything in regards to how long it would take to master a bunch of skills and spells.

Thanks Cerunnir, you wrote that better than I was going to.

I would add that I think a system like that would make practicing even more boring. With how it is now, I at least get the tiniest amount of gambling thrill when I decide to practice my skills. If I knew I had to type 'c vortex;c vortex;c vortex;c vortex;c vortex;sleep" 200 more times before it was perfected I'd probably never play an invoker again.

n/t

In my world there is no such thing as luck, only chance. Everything evens out eventually.

That by itself adds enough complexity to increase per-character variance quite a bit. I simulated your spell mastery scenario as coin flips for 35 characters for 1000 casts (M=500.17 increases, SD=15.83). Then I simulated the same scenario but where the coin flip only takes place on spell failure (assuming the invoker would prac up to 75% for spells he needs to master, and that such a skill succeeds 75% of the time), again for 1000 casts (M=40.83 increases, SD=8.77).

I'll define lucky characters as characters that learn at a rate 2 standard deviations above the mean and unlucky characters as characters who learn at a rate 2 standard deviations below the mean.

In the coin-flip simulation, lucky characters got about 1.1 times as many increases as unlucky characters. But in the flip-on-failure simulation, lucky characters got about 2.5 times as many increases as unlucky characters. (And if we consider 3 standard deviations out, or the luckiest and unluckiest out of a hundred characters, then the luckiest character gets 4.6 times as many increases as the unluckiest character.)

I wouldn't want to generalize these results to the reality of CF, but the comparison does show why simply asserting the law of large numbers isn't good enough. There is still going to be variance, and in complex systems it can be quite meaningful. (Not to mention that the law of large numbers only promises a predictable future; it doesn't counteract a crappy past. So from a subjective standpoint, it doesn't do much for a player's motivation.)

Really, the only reason to have randomized skill increases is for behavioral conditioning. But the behavior it conditions most directly is skill spamming, not playing the intended game. OTOH, it seems a little too late to change that aspect of the game; for me, the law of unintended consequences is convincing where the law of large numbers is not.